ELSA comprises a nationally representative sample of adults aged 50 years and above living in private households in England.  The Study is well placed to conduct research on issues surrounding social isolation and health, as it collects longitudinal information on different components of social isolation and engagement, including marital status, frequency of contact with family and friends (via face-to-face meet-ups, telephone, email/writing, and text messages), participation in various groups and organisations, and number of people living in the household. Coupled with rich longitudinal data on a large array of health conditions, outcomes, and behaviours, ELSA is incredibly useful for researchers wishing to explore the impacts of social isolation. Research using ELSA data has already shown that,  compared with older adults who are not socially isolated, those who are isolated tend to be a higher risk of early mortality, cardiovascular disease, cognitive decline and dementia, increased inflammation, frailty, and depression. Isolated older adults also show poorer health behaviours, including smoking and less physical activity, poorer physical performance, as well as higher rates of falls and hospital admissions.

Social isolation and COVID-19

Recently, I have been working with Professor Andrew Steptoe to better understand the COVID-19 experience of older adults who were already socially isolated prior to the outbreak of the pandemic, with respect to their health, wellbeing, and health behaviours. Were they worse off than those who were socially engaged before the pandemic? Had the pandemic restrictions exacerbated existing health conditions and disparities, for instance due to planned treatments being cancelled or postponed, and the uncertainty and changes experienced throughout the course of the pandemic? Moreover, had they experienced a larger increase in loneliness, with their already limited social activities being further reduced?  

Or, was it the case that those who were already socially isolated before the pandemic actually experienced less of a loss or change in lifestyle when restrictions were imposed, as opposed to their non-isolated counterparts, who experienced a substantial shock to their social system? Could they have had a smaller response in terms of deteriorations in health and wellbeing that were observed in the older adult population during the pandemic? Were they in fact better placed to cope and adapt, potentially via existing routines and arrangements that supported their socially isolated lives?  

We interrogated the ELSA dataset, particularly the COVID-19 data collection in June-July and again in November-December 2020, and the pre-COVID-19 data collection across 2018 and 2019.  We classified participants as socially isolated or not isolated before the pandemic, using an index which included marital status, having limited contact with family and friends, and membership of organisations, clubs, or societies. From the outset, and in line with past evidence, we observed that isolated participants reported lower life satisfaction, quality of life, and health; higher levels of loneliness, depression, and anxiety; more smoking, less physical activity, and lower sleep quality; and greater worries about their future financial situation.

However, using linear mixed models, and adjusting for various demographic and socioeconomic characteristics (e.g., age, gender, education, wealth, neighbourhood deprivation), we found that isolated participants’ response to the pandemic were small with respect to life satisfaction, quality of life, and loneliness. It was, in fact, the non-isolated participants who experienced greater deteriorations in these outcomes, from before to during the pandemic - although isolated participants did show larger declines in how much they smoked and their levels of physical activity, and were more likely to remain worried about their future financial situation. As for physical and mental health and sleep quality, both isolated and non-isolated participants showed similar trends.

What have we learnt?

We can conclude then that, although isolated older adults generally show poorer outcomes than those who are not isolated, they were somewhat protected from the pandemic on some fronts, particularly wellbeing and loneliness. The restrictions imposed during the pandemic had a more deleterious impact on socially engaged older adults, whereas those who were already isolated, experienced fewer changes in their circumstances.

Our implications are two-fold. First, it is vital to continually care for socially isolated older adults, as both before and during the pandemic we observed clear wellbeing, health, and health behaviour disparities between the isolated and non-isolated groups, consistent with existing pre-pandemic evidence. Second, in times of unexpected emergencies and crises, there is a need to pay further attention to older adults experiencing extreme lifestyle changes due to government measures and policy responses.